Groups of computing nodes generally include one “master” node and any number of “non-master” or “subordinate” nodes. The master node may manage resources, such as a distributed state, and may further coordinate activity among the non-master nodes.